Order in the chaos
The call you can defend, in days not weeks.
A grant section, a commercialization report, a BD analysis, a diligence read -- each one is a document that has to survive review by experts who were not in the room. The reviewer changes. The requirement does not.
The shingle reads scientific and business diligence: the novel, unscoped problem -- the one nobody has a template for -- taken on and solved end to end, then set aside because it is solved. So I work from the record rather than the deck. One operator and a fleet of agents, pointed at the published literature and at your question, settling a noisy system into the few facts your decision rests on. No predictions and no price targets -- just what the record establishes, traced to where it says so.
The record
Published reads, every figure sourced
The consolidation nobody priced in
13 sources
biomanufacturing capacity -- SEC filings, IR disclosures, and trial-registry data for Thermo Fisher, Samsung Biologics, and Catalent, 2020-2026
Eight analysts disagree by 2.1x on how big biomanufacturing is. None of them disagree that Samsung Biologics' margin hit 45% last year, that Catalent stopped filing with the SEC, or that nearly 3,700 cell and gene trials are already committed to future capacity. The filings tell a tighter story than the forecasts do.
How biomanufacturing actually works, and what it is worth
89 sources
biomanufacturing -- history, process, product classes, frontier technologies, market size, filings, patents, AI adoption, and the US and EU regulatory backdrop
A living cell that has been talked into making a medicine is a strange kind of factory: variable where a chemical reaction is not, fragile where a pill is not, and licensed as a process rather than as a product. This is the long version -- sixty years of history, every production stage, every product class, the frontier graded honestly against a maturity scale, and what the filings, registries and patent office say the whole thing is worth.
Biomanufacturing's blurry number
25 sources
biomanufacturing -- definition, market size, 2025-2026 trends, CDMO landscape, and US regulatory backdrop
Eight analysts sized the same industry this year and landed between $19 billion and $40.1 billion. The gap is not sloppiness. It is a category whose edges are still being drawn, live, by workforce shortages, an onshoring boom and a regulator that has started promoting the technology it once merely tolerated.
Pipe dreams
5 sources
provenance -- what it takes for a fact to stay true, and what asking the record now costs
A pitch deck said two patents stood issued. The public record said both had expired. What changed is not that somebody finally checked -- it is what checking now costs.
Every sentence true
8 sources
spin in trial reporting, and the failure mode no citation check can catch
A trial write-up can be accurate in every particular and still leave an expert reader with the wrong impression. This is not a theory: someone ran it as a randomised trial, on 300 clinicians, using real abstracts and their own de-spun rewrites. It is the one failure a citation check cannot catch.
From pilot to protagonist
12 sources
agentic AI in drug development, and why provenance is the moat
Pharma is handing AI the decisions, not just the slides. Revisited seven weeks on, with a finding that moves the argument: the machine's characteristic error is no longer inventing sources but quietly leaving things out -- which is harder to catch, and strengthens the case rather than softening it.
The shape agreed in advance
1 source
what a schema is, why shape buys detectability, and what it costs
A schema is written for a reader who cannot ask a follow-up question. That constraint, not tidiness, is why schemas exist -- and it explains why a false sentence in a document is not merely hard to find but structurally invisible.
What a verified read looks like
5 sources
Immunome, Inc. (IMNM) -- varegacestat in desmoid tumor
A hazard ratio of 0.16 is the kind of number that ends an argument. The work is not finding it -- it is establishing that it means what the press release says it means.
Corbus Pharmaceuticals Holdings CRBP
22 sources · 6 databases
nectin-4 ADC · 2L head and neck + cervical cancer
CRB-701 ORR 47.6% in 2L head and neck. Zero peer-reviewed data. Pliant's αvβ8 lead boxes out CRB-913.
Immunome, Inc. IMNM
97 sources · 17 databases
gamma-secretase inhibitor · desmoid tumor
Phase 3 hazard ratio 0.16. NDA filed April 29. ASCO oral on May 30 fills in subgroup detail.
Wave Life Sciences WVE
15 sources · 7 databases
GalNAc-conjugated AIMer · alpha-1 antitrypsin deficiency
GalNAc cleared 11 µM where one LNP didn't. Beam's LNP cleared it harder.
Compass Therapeutics CMPX
52 sources · 14 databases
DLL4×VEGF-A bispecific · biliary tract cancer
Primary hit. Secondary hit. OS missed because 54% of placebo crossed over.
Absci Corporation ABSI
106 sources · 16 databases
AI antibody design · oncology
The AI designs antibodies. The clinic opened eleven months ago.
I also keep score in public: oncology readouts called before the data, then graded against what printed -- the misses included. See the calibration ->
All research ->How it works
01
You hand me the pile
Your documents, your data room, your question -- in whatever shape they are in. I wire it to the live primary record: filings, trials, patents, labels, the literature. Your private pile and the public record end up in one view.
02
The engine verifies
Every figure traces to its source -- verbatim line, durable id, timestamp. A model of a different lineage audits each claim and can only veto it. The weak claim dies in private rather than in your committee.
03
You get something you can defend
Not a document that asserts, a document that cites. I read every gate myself and hand back the handful of facts the decision rests on, with the logic that connects them, and every digit re-checkable.
The FDA's 2025 guidance on AI in regulatory work expects exactly this provenance by construction -- so the output meets a standard that turns ungrounded AI away.
Working together
The front door
A Verified Read
One asset, fixed scope, about ten business days. Every claim traced to the filing, trial, or paper behind it, and dated before the readout it speaks to. When the question is bigger than one asset, the scope is quoted to the question.
Fixed fee, quoted against the scope -- never an hourly rate. You should be able to drop a line item into a budget and forget it, whatever the hours turn out to be.
Who stands behind it
BigBio is Christopher Davis
Twenty-plus years across the gates this work has to clear: bench discovery at NIBR, diagnostics at Metamark, a $1.3B GI franchise at Shire, founder and CEO of a venture that took a 90-minute sterility test from zero to market, and a decade of seed-stage deal reviews with Mass Medical Angels.
Raised off-grid in a log cabin outside Anahim Lake, British Columbia -- no electricity, no running water -- now working at the bleeding edge of compute in Boston.
Point me at the deal you are weighing or the asset you are building.